This research addresses refinement of a theory of speech production that has existed for more than half a century, the source-filter theory. It is used routinely in acoustic phonetics and in the description of voice and speech disorders. It is also used extensively in speech processing to separate sound sources from sound resonators. Recent advances in speech production research, primarily with computer simulation, have shown that the primary sound source in the larynx (glottal airflow) can be highly influenced by acoustic pressures propagating backwards from the vocal tract to the larynx. The feedback interaction can disturb vocal fold vibration, producing pitch jumps, subharmonic frequencies, or even chaos. It can also affect the phonation threshold pressure, making it either more easy or more difficult to self-sustain vocal fold oscillation. The significance of the research is that cerain vocal sound qualities, often labeled as voice disorders, may be part of a normal inventory of human sounds. Specific aims are (1) to determine if vocal intensity can be affected by source-filter interaction (SFI), (2) to discover which vocal tract configurations lead to the greatest interactions, (3) to validate or invalidate a procedure known as inverse-filtering, which allegedly separates the source characteristics from filter characteristics, and (4) to determine if vocal instabilities in the aging male voice are related to SFI. Human volunteers and computer simulation will be used in a structured way to quantify and explain the nature of SFI.